Corporate News

▼

SENSING TECH Debuts at 2026 Taipei International Industrial Automation Exhibition

SENSING TECH Debuts at 2026 Taipei International Industrial Automation Exhibition

Empowering the Era of Physical AI and Robotics with NVIDIA Jetson Thor Platform + GMSL Low-Latency Vision Solutions From August 19 to 22, 2026, the Taipei International Industrial Automation Exhibition – Asia's premier industrial technology event – was held at the Nangang Exhibition Center Halls 1 and 2. Under the theme "Integrate as One. Power Across Asia.," this year's exhibition spotlighted cutting-edge fields including AI, human-robot collaboration, and humanoid and collaborative robotics. At this grand event, SZ SENSING TECH Co., Ltd. (SENSING TECH)made a striking appearance alongside its partners Apo Star Co., Ltd. and NEXCOM, showcasing its AI vision products and solutions. The exhibition highlighted SENSING TECH's latest achievements in robotic vision and multi-platform adaptation. Leveraging its deep "hardware-software integrated" technical expertise, the company demonstrated the core value of intelligent vision in the Physical AI era to the industrial community. Multi-Camera Low-Latency Vision Solution Based on NVIDIA Jetson Thor Platform At the Apo Star booth, the two companies jointly unveiled a multi-camera low-latency vision system solution based on the NVIDIA Jetson Thor platform. This solution deeply integrates SENSING TECH's industry-leading GMSL camera technology with Apo Star's extensive channel experience accumulated in Taiwan's automation market, with a primary focus on L4 autonomous driving, humanoid robotics, and AMR applications. In the live demonstration, multiple GMSL cameras achieved synchronized acquisition and ultra-low-latency processing through the Jetson Thor™ platform, delivering precise and reliable visual inputs for perception fusion. Leveraging Apo Star's years of deep-rooted presence in Taiwan's industrial automation sector and its profound understanding of local customer needs, combined with SENSING TECH's high-performance vision hardware, the collaboration provides system integrators and equipment manufacturers with a turnkey visual perception foundation. Throughout the exhibition, the joint booth attracted numerous professional visitors, generating enthusiastic on-site response. One-Stop Physical AI Deployment Solution and AI Robot Dynamic Balance Control Platform  At the NEXCOM exhibition area, SENSING TECH, as a key vision perception partner, jointly showcased a comprehensive ecosystem solution covering robotic control hardware and software, AI controllers, and critical components. NEXCOM's subsidiary NexCOBOT highlighted its one-stop Physical AI "ready-to-deploy" solution, featuring self-developed control software and hardware, AI controllers, and broad cross-platform chip support. At the critical component level, NEXCOM integrated SENSING TECH's high-performance GMSL cameras and vision perception solutions. Additionally, NEXCOM demonstrated a humanoid robot AI dynamic balance control platform based on the NVIDIA Jetson Thor platform. SENSING TECH's GMSL series cameras and multi-channel vision synchronization solutions were deeply integrated into this platform, providing high-bandwidth, low-latency environmental perception capabilities for humanoid robots, quadruped robots, and various autonomous mobile robots. This significantly reduces the complexity of vision integration for developers and accelerates product commercialization for OEMs. NVIDIA Thor + GMSL: Breaking Through Multi-Sensor Perception Fusion Challenges at the Edge As Physical AI transitions from concept to reality, terminal devices such as autonomous vehicles, humanoid robots, and AMRs are experiencing surging demand for multi-perspective visual perception. However, synchronized acquisition and low-latency transmission of multiple video streams have consistently remained critical technical bottlenecks in edge deployment. To address this challenge, SENSING TECH has achieved deep integration of automotive-grade GMSL chips through the MIPI interface of the NVIDIA Jetson Thor platform, seamlessly bridging low-latency, high-stability transmission capabilities with NVIDIA's computing platforms. By deeply integrating its solutions with high-performance edge AI systems powered by the Jetson Thor platform from Vecow, SENSING TECH has achieved synchronized acquisition and ultra-low-latency processing for up to 16 GMSL cameras. This creates a "ready-to-use" high-performance vision perception foundation for edge-side customers within the NVIDIA Physical AI ecosystem, effectively removing key bottlenecks in multi-sensor fusion deployment. Deeply Integrated with the NVIDIA Ecosystem, Continuously Empowering Physical AI Deployment As an NVIDIA Elite Partner, SENSING TECH possesses deep technical expertise in vision solution adaptation for the NVIDIA Jetson Thor platform. The multi-camera synchronization solution showcased at the exhibition is deeply customized based on the official Jetson Thor development kit, combining the powerful computing capabilities of the Jetson Thor platform with SENSING TECH's proprietary image tuning and synchronization algorithms. This provides global Jetson Thor developers with mature, reusable GMSL camera application references. Since its establishment in 2018, SENSING TECH has steadfastly pursued a "hardware-software integrated" development path, achieving independent control across the entire spectrum – from underlying lens module design to upper-level algorithm acceleration. The company is now deeply involved in building the vision perception unit for NVIDIA's first open-source humanoid robot reference design, Isaac GR00T, delivering a complete AI vision solution for industry-benchmark robotic products – from head-mounted vision to end-effector execution. This fully validates its end-to-end delivery capability from core components to system-level solutions. Vision as the Foundation, Co-Creating the Future of Physical AI From technological breakthroughs to ecosystem co-building, SENSING TECH remains committed to its core mission: "Letting machines see the world better." Looking ahead, the company will continue to deepen its focus on AI vision perception, dedicated to providing perception infrastructure for the Physical AI world. With higher-performance innovative products and in collaboration with more partners, SENSING TECH strives to accelerate the transition of Physical AI from laboratories to real-world production and living scenarios, helping usher in the full arrival of an intelligent world.

SENSING Launches Ultra-Compact High-Reliability GMSL Capture Box: Palm-Sized Imaging Solution

SENSING Launches Ultra-Compact High-Reliability GMSL Capture Box: Palm-Sized Imaging Solution

As AI transformation sweeps across industries, high-precision and reliable image capture has become the critical "eyes" of intelligent systems. Yet traditional solutions often suffer from bulky size, complex deployment, and compatibility issues – especially for specialized equipment like automotive GMSL cameras that require intricate cabling and capture cards, stifling innovation. Breaking Boundaries with Miniaturized Excellence SENSING introduces the PANDA Serial PG2, a rugged ultra-compact GMSL capture box delivering: ✓ Lossless GMSL-to-USB3.0 conversion ✓ Industrial-grade reliability in 45×64×23mm (62% smaller than gen1) ✓ Zero-compromise 1080P@60fps YUV/RAW streaming Automotive-Grade Connectivity Redefined Featuring reinforced USB3.0 connectors certified to vehicle vibration standards (ISO 16750), the PG2 maintains 99.99% connection stability under harsh conditions – solving chronic contact failures in robotics/ADAS testing where conventional ports falter. Pocket-Sized Powerhouse The 62% volume reduction belies its capabilities: <1ms latency for embodied AI real-time response Uncompressed data pipeline preserving original image fidelity Dual-mode power delivery (12V DC or bus-powered via USB3.0) Plug-and-Play Agility With native UVC compliance, the PG2 works driver-free across Windows/Linux. Its coaxial-to-USB direct link eliminates traditional capture cards, cutting deployment time by 70% for lab/field setups. Smart Imaging, Evolving Exclusive to SENSING cameras: In-field firmware upgrades for adaptive imaging profiles Scene-optimized processing (low-light/HDRI/high-speed) Multi-camera sync for 360° data harvesting Cross-Industry Enablement ▸ Embodied Intelligence: Sub-5ms visual feedback loops▸ Mobile Labs: Synchronized multi-camera recording▸ Research: Lab-ready with zero configuration The Future in Focus By condensing pro-grade capture into a palm-sized device, the PANDA PG2 shatters legacy constraints. SENSING continues to pioneer AI perception infrastructure, pushing the boundaries of what machines can see.

SENSING TECH Debuts at WAIC 2026 | From Visual Perception to Data Infrastructure, Full-Stack Capabilities Drive Embodied AI Industrialization

SENSING TECH Set to Debut at 2026 Taipei International Industrial Automation Exhibition

From July 17 to 20, the 2026 World Artificial Intelligence Conference (WAIC) and the High-Level Meeting on Global AI Governance were held at the Shanghai World Expo Exhibition & Convention Center. Under the theme "Intelligent Partners, Creating the Future Together," the exhibition area surpassed 100,000 square meters for the first time, with over 1,100 enterprises showcasing more than 3,000 cutting-edge technological achievements. As a leading enterprise in AI vision solutions, SZ SENSING TECH Co., Ltd. (SENSING TECH) made a prominent appearance with its flagship core products and innovative solutions, delivering an immersive full-stack technology experience—from visual perception to data collection—for on-site visitors. Multi-Platform Embedded Vision Solutions: Building an Open and Compatible Perception Foundation During the exhibition, SENSING TECH comprehensively showcased its core vision products for intelligent equipment and robotics, covering multi-platform embedded vision solutions, GMSL camera solutions, and a variety of innovative imaging products. In the field of multi-platform embedded vision, SENSING TECH demonstrated vision solutions that fully support mainstream AI computing platforms including NVIDIA, Intel, and D-Robotics. As an NVIDIA Elite Partner, SENSING TECH fully leverages the edge-computing capabilities of each platform, delivering high-precision, low-latency vision perception systems through software-hardware co-optimization. Its embedded vision camera series provides developers with official development kits, camera solutions, and ISP tuning services, covering the entire workflow from prototype validation to large-scale deployment. Among these, the embodied AI vision solution based on the NVIDIA Jetson AGX Thor drew significant attention at the event. This solution fully unleashes the powerful computing capabilities of the Thor platform, flexibly supporting up to eight self-developed robot cameras from SENSING TECH. It integrates binocular and RGB-D vision technologies, allowing flexible deployment across critical robot components such as the head and torso. Through high-precision time synchronization and microsecond-level trigger coordination, it builds a full-coverage visual perception network for robots, facilitating precise navigation, real-time obstacle avoidance, operational guidance, and motion feedback—significantly enhancing interaction and execution capabilities in complex environments. Additionally, SENSING TECH showcased its embodied AI vision solution based on the D-Robotics Sunrise S600 platform. This solution combines SENSING TECH's self-developed high-precision GMSL cameras with the S600's high on-device inference capabilities, enabling collaborative processing of multimodal perception and real-time control directly at the chip level—offering another high-performance hardware pathway for autonomous decision-making in embodied intelligent robots. With this addition, SENSING TECH's embedded vision product line now covers a broader spectrum of computing power tiers, from high-power flagship-grade to cost-effective edge-side solutions, enabling more flexible responses to diverse customer requirements regarding cost, power consumption, and performance. Industrial Vision Solutions Driving Intelligent Inspection and Guidance In the field of 2D vision-guided and industrial inspection, SENSING TECH's full-stack vision solution—spanning algorithms to hardware—was another highlight at the booth. This solution integrates core algorithms for image acquisition, processing, and recognition, and seamlessly supports mainstream platforms including Intel, NVIDIA, and Rockchip, allowing flexible computing power selection. It covers positioning, inspection, measurement, barcode reading, and other functions, with the ability to be customized for different production lines and specific targets. The solution is widely applicable to scenarios such as 3C electronics, semiconductors, logistics AGV/AMR, lithium batteries, and industrial patrol inspections. As the core hardware foundation of this solution, the SI series industrial area-scan cameras were also exhibited. Based on the GMSL protocol, this series offers resolutions ranging from 2 to 20 megapixels, supports global/rolling shutters, and delivers clear imaging under complex lighting conditions. With power consumption as low as 0.7W, a single cable handles power supply, data, and transmission, supporting stable transmission over distances up to 15 meters with strong anti-interference capabilities. Featuring a computing-sensing separated design, low temperature drift, and high precision, the cameras ensure stable long-term operation and are well-suited for harsh industrial environments. First-Person Data Collection System Camera Customization Service At this year's WAIC, SENSING TECH unveiled its first-person perspective data collection system camera customization service, which became one of the most talked-about focal points at the booth. As the embodied AI industry transitions from prototype demonstrations to batch deployment, high-quality, multimodal real-world scenario data has emerged as a critical bottleneck constraining robotic intelligence. SENSING TECH's first-person data collection system camera customization service allows operators to perform tasks, perceive, and interact with the physical environment from their own perspective, while synchronously collecting multidimensional data including vision, tactile feedback, and spatial posture. This data collection paradigm aligns more closely with the model's perceptual perspective, helping to enhance the robot's task comprehension and decision-making capabilities. Electro-Optical Gimbals Empowering All-Weather Intelligent Perception for Unmanned Platforms At the exhibition, SENSING TECH also presented its electro-optical gimbal series specifically designed for robotic dog patrol inspections and emergency rescue applications. Among them, the SGU-HS1-L10D integrates a high-pixel wide-angle visible-light camera, an ultra-high-definition optical zoom camera, a high-resolution thermal imager, and a high-precision laser rangefinder, achieving multi-spectral fusion perception. The lightweight three-axis gimbal, coupled with an adaptive shock-absorption module, automatic focusing, and gimbal stabilization algorithms, ensures that the image remains clear and stable even during complex movements such as walking and climbing—providing reliable, all-weather unmanned visual perception solutions for various industries. Centered on Visual Perception, Empowering Embodied AI Industrialization As a technology innovation enterprise rooted in imaging and vision technologies, SENSING TECH has adhered to a "hardware-software integrated" development path since its inception—possessing both core hardware design and manufacturing capabilities and deep expertise in algorithm optimization and scenario adaptation. Looking ahead, SENSING TECH will continue to deepen its visual perception technologies and solutions, providing efficient and reliable foundational support for intelligent robots through its solid product portfolio and full-stack technical capabilities, and helping artificial intelligence permeate every industry and sector.

SENSING TECH Debuts at Automate 2026 | Empowering Robots with “Smart Eyes” and “Precise Hands” Through Visual Perception

SENSING TECH Debuts at Automate 2026 Giving Robots Smart Eyes & Precise Hands

On June 25, Automate 2026 – North America’s largest and most influential automation and robotics event – came to a successful close at the McCormick Place Convention Center in Chicago. This year’s exhibition showcased the world’s most cutting-edge robotics technologies, AI applications, machine vision systems, and industrial automation solutions. As a leading provider of AI vision solutions, SZ SENSING TECH CO., LTD. (SENSING TECH) presented its portfolio of core vision products at Booth #11044 in the North Hall, delivering an immersive technology experience spanning from core vision to physical AI control for on-site visitors. Multi-Platform Embedded Vision Solutions Empowering Intelligent Equipment and Robots Throughout the show, SENSING TECH fully demonstrated its series of core vision products for intelligent equipment and robots, covering multi-platform embedded vision solutions, GMSL camera solutions, and a variety of innovative imaging products. In the field of multi-platform embedded vision, SENSING TECH showcased vision solutions that fully support mainstream AI computing platforms including NVIDIA, Intel, and Qualcomm. As an NVIDIA Elite Partner, SENSING TECH leverages the edge-computing capabilities of each platform through hardware-software co-optimization to build high-precision, low-latency visual perception systems. Its embedded vision camera series provides developers with official development kits, camera solutions, and ISP tuning services, covering the entire workflow from prototype validation to large-scale deployment. Notably, SENSING TECH’s embodied intelligence vision solution based on the NVIDIA Jetson AGX Thor also attracted significant attention at this year’s show. The solution fully unleashes the powerful computing capabilities of the Thor platform, flexibly supporting up to eight channels of SENSING TECH’s self-developed robot cameras. Integrating binocular and RGBD vision technologies, it enables flexible deployment at key locations such as the robot’s head and torso. With high-precision time synchronization and microsecond-level trigger coordination, it builds a full-body visual perception network for robots, facilitating precise navigation, real-time obstacle avoidance, task guidance, and operational feedback, thereby significantly enhancing interaction and execution capabilities in complex environments. Industrial Vision Solutions Driving Smart Inspection and Guidance In addition, SENSING TECH presented its full-stack vision solutions spanning algorithms to hardware for 2D vision-guided and industrial inspection applications – another highlight of the booth. The solution integrates core algorithms for image acquisition, processing, and recognition, and is seamlessly compatible with mainstream platforms such as Intel, NVIDIA, and Rockchip. It supports flexible computing power selection and covers functions including positioning, inspection, measurement, and barcode reading. Tailored optimization can be applied for different production lines and targets, making it widely applicable in scenarios such as 3C electronics, semiconductors, logistics AGV/AMR, lithium batteries, and industrial inspection. As the core hardware foundation of this solution, the SI series industrial area-scan cameras were also on display. Based on the GMSL protocol, the series offers resolutions from 2 to 20 megapixels, supports global/rolling shutters, and delivers clear imaging under complex lighting conditions. With power consumption as low as 0.7W, a single cable handles power, data, and transmission, supporting stable transmission up to 15 meters with strong anti-interference capability. Featuring a computation-sensing separation design, low temperature drift, and high precision, the cameras ensure long-term stable operation in harsh industrial environments. Joint Demonstration with NexCOBOT: “Physical AI Control Starts at the Core” Another highlight at the SENSING TECH booth was the joint application demonstration with ecosystem partner NexCOBOT, themed “Physical AI Control Starts at the Core.” The demonstration fully presented a closed-loop technology from core visual perception to physical AI control: SENSING TECH’s high-performance GMSL cameras delivered low-latency, high-bandwidth image transmission under complex lighting and high-speed scenarios; NexCOBOT, based on the NVIDIA AI computing platform, integrated microsecond-level precision motion control with international standard functional safety systems. The joint solution vividly illustrated how visual perception and motion control work in synergy to empower embodied intelligent robots, enabling a seamless transition from “seeing” to “acting.” Seeing a New Future of Automation with “Smart Eyes” Since its founding in 2018, SENSING TECH has consistently focused on imaging and vision technologies, specializing in robotics, autonomous driving, low-altitude economy, and AIoT. The company is committed to providing high-performance vision solutions for its customers, playing a core role in key applications such as embodied intelligence, low-speed autonomous driving, and drones. With its full-stack imaging technology capabilities, hardware-software synergy deeply adapted to mainstream AI chip platforms, and large-scale high-end manufacturing capacity, SENSING TECH has established a leading technological and product position in the industry. Its appearance at Automate 2026 marks a critical step for SENSING TECH in deepening its presence in North America and advancing its global strategy. Sharing the stage with the world’s top companies not only demonstrated the technological strength of Chinese smart manufacturing in visual perception but also conveyed the vision of “Letting Machines See the World.” Looking ahead, SENSING TECH will continue to deepen its focus on visual perception, partner with industry peers with an open attitude, and jointly drive innovation in the intelligent robotics industry.

Empower Robots to "Understand" the Complex World | SENSING TECH Robotics Releases Breakthrough Stereo Depth Algorithm

SENSING TECH to Debut at 2026 Taipei International Industrial Automation Exhibition

SZ SENSING TECH Co., Ltd. (SENSING TECH) has officially launched its Stereo depth algorithm, sDepth. This cutting-edge neural network technology redefines stereo depth estimation, delivering highfidelity, highly reliable depth perception for robotics, industrial automation, AR/3D reconstruction, and beyond. The simultaneous release of the sDepth Software Development Kit (SDK) significantly lowers the barrier to integrating Stereo depth algorithms, helping more intelligent devices truly gain the ability to "understand the complex world." △ sDepth multiscenario demo Why reinventing Stereo depth perception? Binocular stereo vision mimics human binocular ranging and offers inherent advantages for 3D perception. However, traditional algorithms rely heavily on texture features for matching. Faced with white walls, smooth floors, transparent glass, or low‑light environments, matching becomes difficult and depth output degrades significantly. SENSING TECH's stereo depth algorithm (sDepth) uses neural networks to generate high‑fidelity depth maps directly from stereo image pairs, providing reliable depth estimation even under challenging conditions. Compared to conventional methods, it achieves substantial accuracy improvements in texture‑poor and dimly lit areas, making it an ideal choice for autonomous robots, augmented reality (AR), and other applications. Core Technology of sDepth: Neural‑Network‑Driven Perception Innovation The competitive edge of sDepth lies in its end‑to‑end neural network architecture for stereo depth estimation, which combines feature learning with multi‑strategy matching.   End‑to‑End Neural Network Architecture Generates high‑fidelity depth maps directly from stereo image pairs, effectively solving matching problems on textureless surfaces, repetitive textures, and under challenging lighting. Neural Fusion Architecture Compensates for missing information via deep learning, achieving significant breakthroughs in accuracy for occluded and texture‑less regions. Multi‑Strategy Fusion Mechanism Dynamically integrates multiple matching strategies to ensure output consistency in complex environments. Ultra‑Low Latency / Real‑Time Performance End‑to‑end latency below 5 ms, perfectly supporting mainstream AI computing platforms such as NVIDIA Jetson AGX Orin and Thor. sDepth SDK: Easier Depth Integration To help customers quickly implement binocular depth projects, Sensyn Robotics has developed the sDepth Software Development Kit (SDK). It integrates core capabilities such as calibration parameter loading and depth map generation, covering both offline testing and real‑time capture scenarios, thereby drastically reducing the development threshold. △ NVIDIA Jetson AGX Orin + S36 camera – indoor real‑time effect at 24 FPS Key Features of the sDepth SDK Offline & Real‑Time Dual Modes –Supports offline image tests and live camera data streams, enabling the full workflow from algorithm validation to product deployment. Plug‑and‑Play Calibration Parameters – Load calibration parameters via configuration files, no hard‑coding required. Multiple Display Modes – Switch display modes using keyboard shortcuts without affecting accuracy (visual changes only). Multi‑Platform Optimization – Deeply optimized for platforms such as Jetson AGX Orin and Thor. Broad Application Scenarios   The high‑fidelity depth perception of sDepth is unlocking value across multiple cutting‑edge fields: Autonomous Robot Navigation – Service, cleaning, and inspection robots obtain reliable depth maps on low‑texture floors and under changing lighting, enabling stable obstacle avoidance and smooth navigation. AGV Navigation & Warehouse Automation – In dark or low‑texture warehouses, sDepth outputs continuous and reliable depth information, helping AGVs identify traversable areas and avoid collisions. Industrial Automation – Robotic grasping, workpiece positioning, and assembly guidance – even on reflective or texture‑sparse objects, sDepth delivers stable depth data for sub‑millimeter precision. Augmented Reality (AR) – High‑quality, low‑latency depth maps allow virtual objects to blend precisely with the real environment, enhancing immersive AR experiences. Open Empowerment: Giving Robots "Smart Eyes" Amid the rising wave of embodied AI and spatial intelligence, breakthroughs in visual perception are a critical step toward enabling devices to truly "understand" the physical world. Binocular depth vision is rapidly becoming the core sensing modality for robots, autonomous driving, and industrial automation. SENSING TECH Robotics' sDepth algorithm breaks through the traditional limitations of binocular stereo in low‑texture and low‑light scenarios. With its ultra‑low latency and flexible SDK deployment, it provides a solid vision foundation for the intelligent upgrade of multiple industries. Going forward, SENSING TECH Robotics will continue to iterate the sDepth algorithm and SDK, support more computing platforms and depth modes, and work with industry partners to drive large‑scale adoption of stereo vision in the era of embodied intelligence.

SENSING TECH Makes a Strong Debut at Computex 2026 & GTC Taipei

SENSING TECH Makes a Strong Debut at Computex 2026 & GTC Taipei

TAIPEI, June 2, 2026 – The global technology community has turned its eyes to Taipei. As Asia's largest and the world’s second-largest computer technology exhibition, COMPUTEX TAIPEI 2026 opened under the theme “AI Together.” The event focused on three core pillars: AI Computing, Robotics & Smart Mobility, and NextGen Technology, bringing together 1,500 exhibitors and over 6,000 booths – a record scale. Concurrently, NVIDIA GTC Taipei was held from June 1 to 4 at the Taipei International Convention Center (TICC), gathering developers, researchers, and industry leaders to explore AI Factories, Physical AI, and Robotics. At this year’s technology showcase, SZ SENSING TECH Co., Ltd. (SENSING TECH) – an industryleading provider of AI vision solutions – made a remarkable appearance alongside multiple ecosystem partners. The company demonstrated its latest AI vision products and solutions, highlighting recent advances in robotic vision and multiplatform adaptation. With deep expertise in integrated hardware and software, SENSING TECH showcased the boundless potential of intelligent vision in the era of Physical AI. Empowering the Isaac GR00T Humanoid Robot Reference Design – Injecting “Smart Eyes” into the H2 Plus At GTC Taipei, NVIDIA CEO Jensen Huang officially launched Isaac GR00T, the world’s first open humanoid robot reference design, integrating the Unitree H2 Plus robot, the NVIDIA Jetson Thor platform, and the GR00T software stack. As an NVIDIA Elite Partner, SENSING TECH played a key role in building the visual perception unit of this reference design. It developed a fullseries AI vision solution for the H2 Plus – from head to hands – including a bionic binocular camera for the head and wristmounted operation cameras, complemented by vision software and platform adaptation to deliver a true hardwaresoftware integrated solution. Standing approximately 1.82 m tall, weighing 68 kg, and featuring 31 degrees of freedom, the Unitree H2 Plus can perform complex movements such as ballet and martial arts. Leveraging the powerful computing capabilities of the Jetson Thor platform, SENSING TECH’s solution significantly enhances the robot’s multimodal perception, realtime inference, and autonomous decisionmaking, enabling a seamless transition from “seeing clearly” to “fine manipulation.” Joint Launch with TI: MultiChannel FPDLink Vision Solution Based on Jetson Thor At the Texas Instruments (TI) booth, SENSING TECH – as a TI ecosystem partner – spotlighted its multichannel FPDLink camera AI vision solution based on the NVIDIA Jetson Thor platform. The solution demonstrates SENSING TECH’s flexibility in highspeed image acquisition and transmission for complex scenarios, further highlighting its crossplatform vision integration capabilities. Broad Adaptation with NVIDIA’s Core Ecosystem Partners – MultiPlatform AI Vision Solutions Shine Throughout Computex and GTC Taipei, SENSING TECH’s GMSL series cameras and embedded vision solutions were deeply integrated into the AI vision offerings of several of NVIDIA’s core ecosystem partners. Visitors experienced the following applications at partner booths: Advantech In its edgecomputingenabled AMR intelligent upgrade solution, Advantech adopted SENSING TECH’s multichannel vision solution to provide precise edge perception for autonomous mobile robots (AMRs). AverMedia AverMedia showcased a fourchannel vision solution based on SENSING TECH’s 3MP GMSL2 cameras, using the company’s fourchannel adapter board to achieve synchronized capture and processing of multiple highdefinition video streams. ADLINK ADLINK demonstrated a complete vision solution integrating SENSING TECH cameras. Featuring multicamera realtime synchronized perception and precise timesynchronization architecture, the platform efficiently performs obstacle detection and recognition, reduces system integration complexity, and provides a reliable vision foundation for edgebased AI navigation and motion control. AAEON AAEON demonstrated an intelligent robotic arm automated sorting system. Its smartfactory AI sorting solution uses SENSING TECH’s multichannel vision solution to achieve highly efficient and accurate automated sorting. Continuous Deep Cultivation – Exploring New Frontiers in AI Vision As a technology innovation company focused on imaging and vision, SENSING TECH has grown into a nationallevel “Little Giant” specialized and sophisticated enterprise since its founding in 2018. The company focuses on three core sectors: intelligent driving, panrobotics, and lowaltitude economy, supported by a professional R&D team of 150 people. From core hardware design to algorithm optimization and scenario adaptation, SENSING Tech remains committed to an integrated hardwaresoftware development path, enabling machines to better “see” and understand the world. During Computex 2026 & GTC Taipei, SENSING TECH’s comprehensive presence at ecosystem partner booths not only demonstrated its deep technical expertise in robotic vision to the global industry but also fully embodied its philosophy of open, winwin ecosystem collaboration. Looking ahead, SENSING TECH will continue to deepen its visual perception technologies and solutions, join hands with global partners with an even more open attitude, iteratively evolve mainstream industry solutions, and contribute “Eyes of Intelligence” to the era of Physical AI.  

SENSING TECH Robotics Vision Solution Empowers Comprehensive, Perception Upgrade for Isaac GR00T Reference Humanoid Robot

SENSING TECH Vision Solution Empowers Isaac GR00T Humanoid Robot with Perception Upgrade

June 1, 2026 – SENSING TECH (SZ Sensing TECH Co., Ltd.), a leading provider of robotics solutions, announced that it has empowered the Isaac GR00T reference  humanoid robot with a brandnew AI vision solution for Unitree Robotics’ nextgeneration bionic humanoid robot, H2 Plus. The solution is built on NVIDIA’s latest Jetson Thor robotics computing platform, with SENSING TECH deeply integrating its vision technologies to deliver a fullspectrum perception upgrade – from head to hand – enabling more accurate and intelligent environmental interaction for humanoid robots. At the NVIDIA GTC Taipei conference, NVIDIA founder and CEO Jensen Huang officially unveiled the world’s first open humanoid robot reference design – the NVIDIA Isaac GR00T Reference Humanoid Robot. This reference design is built around the Unitree H2 Plus humanoid robot, the NVIDIA Jetson Thor computing platform, and the Isaac GR00T open software stack, aiming to provide a unified and open development foundation for cuttingedge humanoid robotics research. As an NVIDIA Elite Partner, SENSING TECH’s vision solution serves as a key component of the “visual perception” section of this reference design, further strengthening H2 Plus’s closedloop capabilities from “seeing” to “understanding” to “acting”. Full-Range Vision Solution Equips H2 Plus with “Intelligent Eyes” The comprehensive vision solution developed by SENSING TECH for Unitree’s H2 Plus covers the entire visual perception chain from the robot’s head to its arms. As an Elite Partner in the NVIDIA ecosystem, SENSING TECH’s Jetson Thorbased robotics vision solution has been integrated as a core perception component of the reference design. Through multiview sensing – including a bionic stereo head camera and wristmounted operation cameras – SENSING TECH provides H2 Plus with visual perception capabilities covering key nodes across the robot’s body. In addition to hardware, SENSING TECH also supplies matched vision software and fullplatform adaptation. The solution fully integrates SENSING TECH’s deep hardwaresoftware expertise in intelligent vision – covering core hardware design, algorithm optimization, and scenario adaptation – enabling robots to achieve highprecision, lowlatency environmental understanding and autonomous decisionmaking in complex environments. Deep Integration with Open Humanoid Robot Reference Design Accelerates Commercialization NVIDIA’s newly released Isaac GR00T Reference Humanoid Robot design, for the first time, unifies the humanoid “body” (Unitree H2 Plus) and the “brain” (Jetson Thor + Isaac GR00T software) into an open, unified development platform. Research teams can directly perform the entire workflow – from data collection and simulation training to realworld deployment – on this platform without reinventing underlying infrastructure. Unitree H2, released in October 2025, is a nextgeneration bionic humanoid robot standing 180 cm tall and weighing 70 kg. It features more than 31 degrees of freedom across its body, with a maximum joint torque of 360 N·m. H2 is capable of performing flexible movements such as ballet turns, backbends, and leg lifts, as well as smoothly delivering martial arts moves like punches, knee raises, and side kicks. The vision solution introduced by SENSING TECH takes H2 Plus’s visual perception capabilities to a new level – leveraging the powerful edgecomputing capability of the NVIDIA Jetson Thor platform, and through hardwaresoftware cooptimization, it significantly improves the robot’s overall performance in multimodal perception, realtime inference, and autonomous decisionmaking. Ultimately, this solution allows the robot to not only “see” its environment clearly but also “manipulate” it with precision, truly achieving seamless integration from perception to control. Deepening Visual Perception, Expanding Robotics Solutions As a technology innovation company focused on imaging and vision, SENSING TECH has from its inception pursued a hardwaresoftware integrated development path, mastering core hardware design and manufacturing while deeply investing in algorithm optimization and scenario adaptation. By continuously advancing visual perception technologies and solutions, and adopting an open approach in partnering with ecosystem allies, SENSING TECH is committed to driving innovation and progress in the intelligent robotics industry – helping machines better “see” the world and more intelligently serve humanity.

SENSING Launches NVIDIA Jetson Thor Visual Solution, Igniting the Revolution in Robotic Visual Perception

SENSING Launches NVIDIA Jetson Thor Visual Solution, Igniting the Revolution in Robotic Visual Perception

On August 25, NVIDIA officially launched its next-generation computing platform, the NVIDIA Jetson AGX Thor developer kit and production-ready modules, specially designed for physical AI. SENSING, with its strong technological capabilities in "imaging and visual technology," has become an NVIDIA Jetson robotics ecosystem partner, working with Jetson ecosystem partners to accelerate the widespread implementation of robotics industry applications. Image Source: NVIDIA Official Website The NVIDIA Jetson Thor is hailed as "the new engine for the physical AI and robotics era," offering unparalleled performance and scalability in a compact, energy-efficient form factor. Powered by the advanced Blackwell GPU and 128 GB of memory, it delivers up to 2070 FP4 TFLOPS of AI computing power, easily running the latest generative AI models—all within a 130W power consumption range. Compared to its predecessor, the Jetson AGX Orin, Thor's AI computing power has increased by up to 7.5 times, and energy efficiency has improved by 3.5 times, providing developers with the massive computing power and sensor support needed for highly complex robotic systems. Image Source: NVIDIA Official Website SENSING, a technology-driven company specializing in imaging and visual technology, possesses full-stack technical capabilities ranging from software-driven development, ISP algorithms, to FPGA algorithm integration. It has built a leading hardware-software collaborative ecosystem in the industry. As early as 2022, SENSING became an NVIDIA Jetson elite partner, providing sensor camera functionality support for the Jetson developer ecosystem. Currently, SENSING has adapted various high-performance camera products to the NVIDIA Jetson Thor platform, offering customers high-quality image solutions. Among these products, the Holoscan Sensor Bridge (HSB) camera solution is the core highlight of SENSING's adaptation to the NVIDIA Jetson Thor platform. With the HSB, SENSING enables the ultra-low latency direct transmission of sensor data from cameras, radar, LiDAR, and other devices to the GPU memory of Jetson Thor, leading the way in high-performance visual solutions adapted to NVIDIA Jetson Thor. This revolutionary development will bring a breakthrough in robotic visual perception. The SENSING HSB camera solution is specifically designed to meet the "high resolution, ultra-low latency, and high computing power demands" of complex robotic application scenarios. The solution offers flexible multi-sensor expansion capabilities, supporting a variety of sensors such as LiDAR, cameras, and Inertial Measurement Units (IMU), providing developers with plug-and-play solutions. SENSING not only provides a complete HSB developer kit but also offers deep customization services based on specific customer needs to help build advanced perception systems. In addition to the HSB camera solution, SENSING also offers more camera products compatible with the NVIDIA Jetson Thor: GMSL Series Cameras Sensing’s GMSL cameras are widely used in scenarios requiring long-distance, low-latency, high-quality transmission. These cameras are compatible with the NVIDIA Jetson AGX Thor platform and have been fine-tuned for image quality, enabling customers to use them directly, significantly shortening development time. MIPI Series Cameras Sensing’s MIPI cameras are characterized by high optical accuracy, reliability, and good temperature drift performance. These cameras are suitable for a variety of short-distance transmission scenarios such as robots, handheld devices, cleaning machines, and AGVs. These cameras are also compatible with the NVIDIA Jetson AGX Thor platform and have been pre-tuned for image quality, offering customers ready-to-use solutions that drastically reduce development time. SENSING’s camera products will continue to evolve towards smaller sizes, stronger performance, and smarter features to meet customer customization and varying scene requirements. SENSING's visual solution based on NVIDIA Jetson Thor can be applied in embodied robot visual perception, industrial inspection, autonomous driving, and other fields, helping customers efficiently develop systems and accelerating the implementation of robotics industry applications. By combining NVIDIA Jetson Thor’s top-tier computing power platform with SENSING’s visual technology, a revolutionary breakthrough in robotic visual perception has been achieved. As AI technology continues to advance, SENSING will deepen its collaboration with NVIDIA to provide higher-quality, more efficient visual solutions for complex AI applications such as robotics and autonomous driving, driving the entire industry toward an intelligent era.

SENSING TECH’s Latest 8MP Ultra-Sensitive Light Camera: Breaking New Ground in Front-View Perception with Superior HDR & LFM Performance

SENSING TECH Set to Debut at 2026 Taipei International Industrial Automation Exhibition

In the current era of rapid iteration in advanced driver assistance systems (ADAS) and autonomous driving technologies, visual perception systems are facing unprecedented challenges — from intense glare and backlighting to low-light night conditions, from LED traffic signal flicker to stringent synchronization requirements for multi-sensor fusion. Every detail tests the extreme performance of automotive cameras. Recently, SENSING TECH launched its latest 8MP automotive-grade camera module, the SG8-OX08DC-5300-G2G-Hxxx. Leveraging top-tier core hardware, professional image tuning, and automotive-grade protection design, this product offers an ideal solution that combines high performance and high reliability for intelligent driving front-view perception. Top-Tier Core Hardware for Exceptional Image Quality The image sensor in SENSING TECH’s SG8-OX08DC series camera adopts OMNIVISION’s first 8MP CMOS image sensor, the OX08D10, featuring 2.1µm TheiaCel™ technology. This technology integrates next-generation Lateral Overflow Integration Capacitor (LOFIC) with OMNIVISION’s DCG™ high dynamic range technology, achieving breakthrough improvements in both dynamic range and LED flicker mitigation (LFM). It accurately captures LED lights under various driving conditions, completely eliminating flickering artifacts. Paired with the indie GW5300 image signal processor and the Maxim (ADI) GMSL2 serializer MAX96717G, the SG8-OX08DC series supports uncompressed YUV data output, preserving raw image details without loss while meeting the stringent automotive requirements for long-distance, high-bandwidth transmission. Three Key Advantages Addressing ADAS Front-View Pain Points Superior HDR Performance – Effortless Handling of Complex LightingThanks to TheiaCel™ technology, the OX08D10 captures high dynamic range images within a 200m range, achieving an optimal balance between SNR1 and dynamic range. Combined with SENSING TECH’s professional image tuning, the SG8-OX08DC series retains rich detail in both highlight and shadow areas even in extreme lighting scenarios such as tunnel entrances/exits and nighttime high-beam exposure, significantly improving the perception reliability of ADAS front-view vision. Leading LFM Performance – Complete Elimination of LED Flickering IssuesConventional rolling shutter sensors are prone to LED “black-out” or flickering when facing traffic signals, creating safety hazards. The OX08D10 leverages the synergy between LOFIC and DCG™ technology to deliver industry-leading LED flicker mitigation (LFM). This enables the SG8-OX08DC series to stably capture the true state of LED light sources under various ambient lighting conditions, providing precise traffic signal perception for urban road scenarios. Excellent Low-Light Imaging – All-Weather Perception QualityIn low-light environments such as nighttime, tunnels, and underground garages, the SG8-OX08DC series maintains industry-leading low-light image quality, ensuring that vision algorithms can still perform effective object detection under insufficient illumination, thereby extending the all-weather applicability of intelligent driving systems. Automotive-Grade Design – Ready for Front-View ChallengesDesigned as an automotive-grade module for harsh environments, the SG8-OX08DC series excels in hardware reliability: Automotive-Grade Protection: The entire module achieves an IP67 protection rating, equipped with AA automotive-grade lenses, offering excellent dust and water resistance suitable for various complex driving environments. Multi-Camera Synchronization: Supports external trigger and multi-camera synchronization, meeting the stringent time synchronization requirements of multi-sensor fusion systems and simplifying the development and integration of front-view multi-camera vision systems. Multiple FOV Options: Available in various field-of-view specifications to flexibly match different front-view perception needs. Deep Compatibility to Accelerate Thor IntegrationThe SG8-OX08DC series has already been integrated into the NVIDIA Drive Thor platform. Leveraging the high AI computing power of the Thor platform, it fully unleashes the high-resolution advantages of the 8MP image sensor, providing high-quality, low-latency image data input for advanced intelligent driving algorithms. This deep hardware-software synergy offers an efficient and reliable front-view perception reference design for intelligent driving systems, significantly shortening system integration and algorithm deployment cycles. Wide Applicability – Empowering Next-Generation Intelligent SystemsWith its high performance and high reliability, the SENSING TECH SG8-OX08DC series is specifically designed for front-view exterior applications and can be widely used in front-view ADAS and vision fusion systems, high dynamic front-view perception, and more. As autonomous driving levels advance and sensor fusion technologies mature, high-quality, high-dynamic, and highly reliable front-view image perception is becoming a decisive factor for system safety. The launch of this camera module series provides the industry with a mature front-view vision product that truly copes with complex road scenarios and delivers all-weather, precise perception. SENSING TECH – Illuminating the Future of Intelligent Driving with New Vision!

SENSING TECH at NVIDIA GTC 2026: Partnering to Shape the New Blueprint for Robotic Vision

SENSING TECH at NVIDIA GTC 2026: Partnering to Shape the New Blueprint for Robotic Vision

The global premier event in artificial intelligence and accelerated computing, NVIDIA GTC 2026, recently concluded successfully in San Jose, USA. As a leading provider of intelligent vision solutions, SENSING TECH made a significant impact at the conference, showcasing its cutting-edge technological advancements. Through keynote presentations and a collaborative exhibition area, the company presented its innovative vision solutions based on the NVIDIA Jetson Thor platform and their deep applications in robotics, becoming one of the highlights of the event. Keynote Speech: Robot Vision Solutions Based on Jetson Thor At a dedicated forum during GTC 2026, Li Bin, CMO of SENSING TECH, delivered an insightful keynote titled "Robot Vision Solutions Based on NVIDIA Jetson Thor." He delved into the challenges of computational power and real-time performance that robotic vision faces in complex scenarios. He elaborated on how SENSING TECH leverages the powerful edge computing capabilities of the Jetson platform, combined with software-hardware co-optimization, to create high-precision, low-latency visual perception systems. Li Bin also noted that SENSING TECH's robot vision solutions based on Jetson have achieved large-scale deployment in fields such as embodied robotics, intelligent manufacturing, and autonomous driving. These solutions significantly enhance a robot's environmental understanding and autonomous decision-making capabilities. His forward-looking technical insights and rich implementation cases sparked lively discussion and garnered high recognition from the audience. Collaborating with Ecosystem Partners to Chart a New Course for Robotic Vision Deployment Beyond the compelling keynote, SENSING TECH joined forces with its partners Vecow, NexCom & NexCobot, and SEEED to create a joint robot vision exhibition area. This collaboration vividly embodied the ecosystem philosophy of "open cooperation for a shared future." Vecow and SENSING TECH co-exhibited a 16-channel vision deep fusion solution based on the NVIDIA Jetson Thor platform. Leveraging powerful parallel processing capabilities and industrial-grade reliability, this solution provides robust support for multi-channel vision applications in demanding environments like intelligent manufacturing and security surveillance. At the NexCom & NexCobot booth, a live demonstration featured a humanoid robot equipped with SENSING TECH's multi-channel cameras performing real-time grasping tasks. Enabled by precise vision-guided algorithms, the robot's movements were smooth and its positioning accurate, showcasing the immense potential of vision technology in flexible manufacturing and human-robot collaboration. SEEED showcased the deep integration of its Jetson-based edge computing kit with SENSING TECH's vision solutions. This combination offers developers a one-stop platform, facilitating everything from prototype validation to rapid deployment and significantly lowering the barrier to entry for robotic vision applications. Spanning core components to complete systems, and algorithmic models to practical applications, the joint exhibition area comprehensively demonstrated the technical strength and innovative achievements of SENSING TECH and its partners in the field of robotic vision, drawing numerous professionals to engage and exchange ideas. Helping Machines "See" the World to Serve a Better Future Being invited to participate in GTC 2026 represents not only NVIDIA's strong recognition of SENSING TECH's technical prowess in robotic visual perception but also serves as a profound journey of technological foresight and industry insight for SENSING TECH. We will leverage this opportunity to empower our customers comprehensively with even more refined technology and superior service. Moving forward, SENSING TECH will continue to deepen its expertise in visual perception technology and solutions. With a more open approach, we will collaborate with a growing number of partners to drive innovation and progress in the intelligent robotics industry, helping machines "see" the world to serve a better future!

High Value, High Precision: SENSING Visual Solutions Empower L4 Autonomous Driving

High Value, High Precision: SENSING Visual Solutions Empower L4 Autonomous Driving

With the accelerated deployment of L4 autonomous driving in low-speed scenarios like unmanned delivery and logistics, visual perception systems are facing higher demands. SENSING recently launched its new L4 autonomous driving vision solution, achieving comprehensive upgrades across four core dimensions: miniaturization, cost-effectiveness, high sensitivity, and all-environment compatibility, providing the industry with more reliable and efficient perception capabilities. Upgraded Miniaturization & Cost-Effectiveness: Enabling Flexible Deployment in Confined Spaces Addressing the stringent installation space constraints of unmanned vehicles, SENSING has not only completed a full miniaturization upgrade of its perception cameras but also achieved a dual breakthrough in performance enhancement and cost reduction. After two generations of upgrades, the perception camera dimensions have been optimized from 30×30×22.5mm to 25×25×19mm, with a newly introduced ultra-thin model boasting a thickness of only 13mm. The cumulative volume reduction exceeds 58%, significantly enhancing adaptability to tight spaces. While achieving miniaturization, the products also deliver improved performance and optimized costs, creating vision solutions that are smaller, more powerful, and more cost-effective for customers. To meet the cargo compartment monitoring needs of unmanned logistics vehicles, SENSING introduces the SG1Z2AESH cargo compartment monitoring camera. Equipped with a high-performance sensor and a fisheye lens, it achieves panoramic coverage inside the compartment at a highly competitive cost. Designed specifically for the interior environments of logistics vehicles, this camera can monitor cargo status, loading rates, and anomalies in real-time, providing customers with a super cost-effective vision solution. Imaging Performance Leaps Forward: 8MP Low-Light & HDR & Superior LFM for Traffic Light Suppression Traffic light recognition is a core challenge in urban autonomous driving scenarios. SENSING's recently released new-generation traffic light recognition camera features excellent low-light imaging capability, supports 120dB High Dynamic Range, and incorporates LED Flicker Mitigation technology. It can handle backlighting, strong glare, and complex nighttime lighting conditions to accurately identify traffic lights and signs. Furthermore, this camera features ultra-low latency transmission and multi-camera synchronous triggering, meeting the high real-time and synchronization requirements of L4 algorithms. All-Weather Application Upgrade: Stable Performance Around the Clock, Unlocking More Scenarios To ensure the stable operation of unmanned vehicles in extreme weather conditions such as rain, snow, fog, and ice, SENSING introduces the industry-leading RedFox-D3Gx camera with automatic defogging and de-icing capabilities. This camera integrates a Sony CMOS sensor, a built-in image processor, and an ADI GMSL serializer. It features electronic lens heating, anti-glare, low latency, and multi-camera synchronous triggering. It complies with automotive-grade performance standards, has undergone professional image tuning, is paired with an automotive-grade lens, and boasts an IP67 high protection rating, enabling it to output stable, clear, high-quality images in various harsh environments. The RedFox-D3Gx camera adopts a POC power supply architecture, supports GMSL interfaces and coaxial power supply, allowing seamless compatibility with existing domain control systems for easy upgrades without hardware changes. The camera's automatic heating delivers excellent defogging and defrosting results. In environments as cold as -40°C, it can rapidly remove ice and fog within minutes, restoring a clear view. Available with various lens angles like 60°, 120°, and 190°, it meets the selection needs for multiple scenarios such as front-view, side-view, and surround-view, making it suitable for extreme climates, mining areas, cold storage facilities, and other complex environments. Now, let's watch the RedFox-D3Gx camera's defogging, defrosting, and de-icing effects under various extreme test conditions. ☟☟☟ Comprehensively Empowering Multi-Scenario Deployment of L4 Autonomous Driving Through continuous product iteration, SENSING has strengthened its technological advantage in the field of autonomous driving visual perception. In the future, the company will continue to focus on customer scenario requirements, promoting the evolution of vision solutions towards greater efficiency, reliability, and economy, facilitating the large-scale commercial application of L4 autonomous driving.

SENSING and Advantech Lead GMSL Camera Integration on NVIDIA Jetson Thor Platform

SENSING and Advantech Lead GMSL Camera Integration on NVIDIA Jetson Thor Platform

Advantech, a global leader in industrial edge computing and AI systems, together with SENSING TECH, a global leader in AI vision solutions, announced they are one of the first to complete the successful integration of the GMSL camera into NVIDIA Jetson Thor-based system MIC742AT, establishing a new leadership position in robotics vision. This milestone underscores the companies’ joint leadership in delivering high-performance vision systems for robotics, autonomous mobile robots (AMRs) and physical AI applications. Why This Breakthrough Matters 1. Optimized for Physical AI, Robotics and AMRs The NVIDIA Jetson Thor module is purpose-built for “physical AI” workloads, applications where the compute unit is embedded in a machine, interacting in real-time with physical environment. These use-cases demand ultra-low latency, very high data bandwidth vision sensor input. By equipping the MIC-742-AT with SENSING TECH GMSL camera, Advantech and SENSING TECH deliver a vision-input subsystem capable of meeting those demands: high-speed GMSL2 link, automotive-grade robustness, and high bandwidth for multi-camera configurations. 2. End-to-End Integration Leadership This achievement goes far beyond a basic “camera compatibility” announcement. Advantech and SENSING jointly completed full hardware signal mapping, serializer/deserializer integration, device-tree and driver validation, and Jetson Thor BSP compatibility to ensure the GMSL2 camera chain works reliably on MIC-742-AT. As first-movers in this integration, Advantech and SENSING TECH provide customers with a pre-validated, production-ready vision-AI system, reducing development risks and accelerating time-to-deployment for robotics and edge-AI platforms. “This GMSL2 integration on MIC-742-AT strengthens our leadership in physical AI. By partnering with SENSING TECH, we now offer a high-performance, low-latency vision platform ready for next-generation robotics.” said Magic Pao, Associate Vice President at Advantech. “SENSING TECH delivers low-latency, high-stability AI vision solutions, offering NVIDIA Jetson Thor users a turnkey experience that truly empowers customers to build proven and reliable vision systems.” Market impact & competitive advantage   Through this collaboration, Advantech and SENSING TECH deliver a differentiated solution that strengthens the robotics and intelligent mobile systems ecosystem: · Accelerated Time-to-Market: Customers can deploy a complete, validated sensor-to-compute platform without navigating the complexities of GMSL camera integration on their own. · High-Performance Visual Computing: Direct GMSL2 connectivity ensures stable signal integrity, long-distance transmission, and minimal latency, ideal for AMR fleets, humanoid robots, and industrial mobility platforms. · Ecosystem-Ready and Scalable: Built on NVIDIA Jetson Thor architecture with a validated GMSL2 pipeline, MIC-742-AT supports multi-camera configurations and offers long-term platform stability. · Reliable Autonomous Perception: Designed for robotics and physical AI, the system enables real-time object detection, environmental understanding, and enhanced autonomous navigation. Driving Next-Level AI Efficiency with MIC-742-AT Powered by NVIDIA Jetson Thor Advantech’s MIC-742-AT, powered by NVIDIA Jetson Thor accelerated computing, delivers exceptional inference performance for advanced perception, sensor fusion, and robot intelligence. When combined with SENSING TECH’s automotive-grade GMSL camera technology, the platform provides robust, high-bandwidth visual processing, essential for mission-critical edge-AI deployments. This integration empowers customers to build reliable, scalable, and production-grade AI systems for smart robotics, industrial automation, and intelligent mobility. Availability The MIC-742-AT, powered by NVIDIA Jetson Thor, and GMSL camera model SG3S-ISX031C-GMSL2F-Hxxx are now available. Learn more at Advantech’s official product page. About SENSING TECH: Sensing Tech specializes in high-quality imaging solutions for AI applications, including intelligent driving/ADAS, automation, robotics, and AIoT. With a strong R&D team and in-vehicle volume production capabilities, many years of imaging expertise to deliver comprehensive camera system solutions, from camera modules to complete imaging systems and ISP tuning.